The use of radioembolization has been shown to be successful for the treatment and downstaging of hepatocellular carcinoma (HCC) as well as liver metastases, and advantages over traditional chemoembolizaion have been reported. The local beta emissions from this technique have been shown to provide therapeutic effect within the tumor. Dosages range from 110-150 Gy, but radiation delivered to malignant tissue is dependent on distance from the yttrium source. Radiotherapy is only effective in treating HCC in doses above 50 Gy. Consequently, treatment response after radioembolization is between 25-60% when based on response criteria in solid tumors (mRECIST). We propose to use ultrasound-triggered microbubble destruction (UTMD) to improve radioembolization of HCC. This technique uses commercial ultrasound contrast agents, whose ultrasound triggered cavitation results in a variety of well documented bioeffects. Localized microbubble cavitation has been shown to sensitize malignant tissue to radiotherapy by inducing vascular endothelial-cell apoptosis. Thus, localized UTMD after radioembolization of HCC may potentially improve tumor response by selectively sensitizing malignant tissue to radiotherapy. In addition, the rate of contrast reperfusion after a destructive pulse can be quantified and used to estimate blood perfusion and fractional vascularity. Changes in fractional vascularity and perfusion may be a potential earlier indicator of treatment response, thereby enabling earlier retreatment in non-responding patients. We propose a clinical trial using localized microbubble cavitation followed by imaging of contrast washing back into the tumor (termed reperfusion) to improve patient outcomes and predict treatment response earlier than the current clinical standards. Patients scheduled for radioembolization of HCC will be randomized to receive radioembolization alone or radioembolization in combination with a contrast ultrasound exam 2-4 hours and 7 and 14 days post radioembolization. A flash destruction-replenishment technique will be used to induce microbubble cavitation at multiple planes within the tumor vasculature. Tumoral response, safety, radioactive bead distribution, liver function, and alpha fetal protein, will then be evaluated and compared between groups. As a secondary aim, tumor perfusion will be quantified in the experimental group based on the rate of contrast agent reperfusion following each destructive ultrasound pulse. Patients will then be stratified according to mRECIST criteria 3-4 months post treatment based on their clinically scheduled follow-up MRI/CT study. Microbubble-derived tumor reperfusion rate will be compared between groups to determine if treatment response can be predicted 7-14 days post therapy. Once validated, this technique is expected to improve patient outcomes by selectively sensitizing malignant tissue to radiotherapy, and by enabling alternative forms of treatment through earlier identification of patients not responding to therapy.